The process and chemistry of converting raw gypsum to a finished commercial board has been developed and refined over many years. Improvements to the process include improved board mixer and kiln designs as well as improvements to the chemistry both defined and non-defined by the process. Improvements have been achieved in terms of board-weight reduction, resistance to water and mould and specialised performance in areas such as fire-resistance and acoustical applications. The current process of wallboard manufacture has been refined to its current state with relatively incremental improvements to the technology with a significant focus on cost improvements exacerbated by the current economic environment.
The Henry Company, a supplier of wax emulsions to the wallboard industry, has completed a significant amount of work in developing and improving additives to the wallboard slurry that can help to improve the economics of the wallboard manufacturing process as well as improve the performance of the finished board. The focus of this work was directed at improving line speed, reducing energy demand and increasing board strength to allow density reduction and/or raw material cost savings. This effort has resulted in proprietary products that allow a reduction in water demand while maintaining the proper rheological profile of the slurry with a concurrent increase in board strength.
Background
The rheology of the stucco slurry mixture is critical to a wallboard manufacturing facility to increase recovery and reduce delay, the two major factors in determining the profitability of a wallboard plant. The viscosity of the mixture and the set time are particularly crucial to give wallboard the consistent quality and ease of use that has endeared it to the construction industry for nearly a century.
With the rapidly increasing cost of energy in recent decades, one focus has been the kiln, where a major amount of money has been spent to improve efficiency and reduce the costs of removing the excess water that is required to make a flowable slurry. The excess water also allows other additives to be easily mixed in, as well as to allow the resulting slurry to be uniformly distributed across the width of the gypsum wallboard line without voids or areas of poor bond to the facing materials. As the amount of water used in the stucco slurry is decreased, although the heat demand and resulting natural gas cost for drying is also decreased, the cost from poor quality board or line shutdown from set board mixers or broken board in the kiln may increase disproportionately. As a result, water reduction in wallboard manufacturing is an area of research for any wallboard plant trying to increase its profit.
During the process of forming the gypsum slurry in the pin mixer of a wallboard line, the consistency of the slurry must be maintained within certain limits to allow proper component mixing, flowout on the forming table and forming plate, dehydration and set, crystal formation and proper drying through the kiln without board defects. The amount of water needed to support the hemihydrate to dihydrate conversion is relatively small versus the amount of water needed to achieve proper mixing and slurry flowout.
One way to reduce the amount of water required for wallboard production is the use of dispersants added to the slurry that allow better wet-out and mixing of components in the board mixer at lower water : stucco ratios. These surface active additives increase the fluidity of the resulting slurry as it is spread across the forming table to fully fill the wallboard surface. Dispersants commonly used can generally be categorised as lignosulphonates, naphthalene sulphonates and polycarboxylate ethers. Each has its advantages and disadvantages in terms of cost and adverse impact with other raw materials in the gypsum wallboard slurry.
The state of play
The wallboard mixer and forming table are critical elements that are key to the wallboard process and represent a complex system of physical and chemical interactions of mixing, flow and stucco set. Traditionally, the wallboard mixer and forming process have been a part of the 'art' of board makers, who by decades of experience have determined the best combinations of machine settings and raw material dosages to give optimum performance on their wallboard lines with their available stuccos.
Computational fluid dynamics has recently been used to attempt to model and understand the wallboard mixer and related equipment. A more direct approach has been to develop lab scale equipment and procedures in order to evaluate dispersants and other additives. This minimises the cost and risk of full-scale plant trials with unproven dispersant or additive technologies. This article uses a rapid bench technique developed in Henry's research and development laboratory in Kimberton, Pennsylvania, USA, coupled with a TA AR2000 rotational viscometer1 to address some of those raw material variables and their interactions with common dispersants, including new proprietary dispersants from Henry Company.
Rheology
With rheology as the primary methodology, a short refresher is in order. Rheology is defined as the study of the flow and deformation of materials. For the purpose of this project, we have focused on flow rather than the deformation characteristics of the stucco slurry. There are two types of flow, shear and extentional. Shear flows are defined as material elements moving over or past each other while extensional flows are defined as material elements that move towards or away from each other. Viscosity is defined as the resistance to all types of flow. Materials are made to flow by imparting a velocity and therefore for a given velocity, the resulting force increases as the viscosity is increased. However, for a given force, the velocity is reduced as the viscosity is increased.
For homogeneous, non-deformable fluids, the equation governing the stress-strain relationships is straight-forward and predictable, but the situation becomes more complicated when the system is a dispersion of particles that are being added and wetted in a continuous aqueous phase within a very short time frame. The mixture is further complicated by additional additives that are used to give the finished wallboard specific performance characteristics.
Further complications arise due to the fact that typical laboratory testing and sample preparation are not conducted on the same time-frame or shear rates as the real world wallboard production. Therefore scaling up can give different results than predicted by lab tests.
This project was focused on the effects of surface active additives in the wet-out of stucco during the short mixing times in a pin mixer, with emphasis on the need to develop an improved laboratory technique to measure this effect in a time period approaching that actually spent in a pin mixer.
Equipment, materials and methods
The experimental design incorporated process-related variables as well as the raw material variables. Water to stucco ratios, premixing procedures and shear rates of the rheometer were examined to determine the best balance of parameters. For example, a range of 100 to 1500s-1 shear rates was narrowed down to a 500s-1 shear rate, which allowed good resolution of the viscosity and set time without spinout between the rheometer discs. Water : stucco ratios in the range of 0.6 to 0.8 were narrowed to a ratio of 0.7 to achieve acceptable mixing in the shortest time while not spinning out on the rheometer. The amount of retarder used allowed measurements to occur on the rheometer without preset, while at the same time minimising the amount that might cause chemical interactions with the other raw materials.
The main raw material variables used were dispersants and stuccos. Three natural stuccos and six synthetic stuccos were examined. They are listed in this study based on the regions defined in the Global Gypsum Directory2. Dispersants examined included a lignosulphonate, a naphthalene sulphonate and Henry's proprietary additives with a variety of formulations.
The laboratory rheometer test procedure mimics the system of the board mixer and the forming table as the slurry is spread across the width of the forming table. A variety of laboratory-scale mixing devices may be used to try to simulate the type and amount of mixing to the slurry as shown in Figure 1, where a traditional laboratory radial/axial flow mixer is shown in the centre and a variety of configurations more closely approximating the board pin mixer are shown on the left. The right-hand image in Figure 1 shows the flexibility of the pin mixing element, with replaceable pins that may be configured in a variety of board mixer styles. Further work will explore other configurations.
The rheometer was then used to measure the properties of the slurry after the board mixer step, as would occur in the setting slurry stucco on the forming table of a wallboard manufacturing line.
The speed with which the slurry moves through the mixer and between the facing sheets is difficult to simulate on a bench scale. As a result, a retarder must be added as is also commonly used in commercial board mixers to maintain the fluidity of the stucco until it can leave the mixing device. The lowest amount of retarder possible is used to minimise any chemical interactions between the retarder and the other additives in the stucco slurry. For example, Figure 2 shows the viscosity change per drop (0.075g) of retarder per lab sample on the right and the effect of retarder on set time on the left.
One of the more traditional methods to assess differences in the fluidity or viscosity of the stucco slurry is a patty diameter measurement made using a Ford Cup as shown in Figure 3 on the left (similar to ASTM C4723) or a slump test as is common with concrete using a two inch diameter tube as shown in the centre. This is similar to the slumps done at some wallboard lines as shown on the right of Figure 3.
The instrument used in this project is an AR2000ex from TA Instruments1, a stress-controlled rheometer capable of conducting rotational, oscillation and rate experiments varying or holding constant shear, strain, temperature or time. The AR2000ex also allows the operator to change the rates of all of the above variables. This versatile piece of equipment provides the ability to analyse both the flow properties of the slurry and the mechanical properties of the board system throughout its evolution from the mixing effects in a pin mixer to full set after the ovens.
The TA AR2000 rheometer allows analysis of the flow properties of the slurry. The slurry components are mixed together and the slurry is then quickly applied to the rheometer for evaluation. Figure 4 shows a typical peak hold rheology scan for eight different additives in simple water/stucco slurries made in the laboratory. During this project, a proprietary mixing protocol was incorporated into the test procedure to allow an extremely accurate scale-up to be accomplished. This minimised the experimental error and had the added benefit of minimising the number of experimental parameter evaluations in field trials.
Introduction of new modifiers
In the industrial environment, a variety of factors is balanced in the successful operation of a board plant. The first step in introducing a new additive to the manufacturing process is to develop a simple bench system to screen additives and dispersants. As the number of dispersant additive candidates was reduced by the screening process, the dispersant additive evaluation expanded to air emissions using a method developed by Henry and then to a commercial scale where the interactions with the process could be observed as well as properties of the final panel such as the dried board core structure.
Stucco is the most critical raw material in wallboard manufacturing. Stucco used at different plants can vary tremendously, not only in colour, as shown in Figure 5, but in other physical and chemical properties. These chemical properties may either enhance the effect of certain dispersants or adversely affect the properties of the final panel.
The experimental design of this project included nine different stuccos (three natural and six synthetic), eight dispersants, three levels of retarder, three different rheometer shear rates and three different water : stucco ratios. Every experiment was replicated at least in triplicate and the results analysed via Minitab statistical software4. The rheometer was run using a stress rate peak hold program that had been developed with the assistance of application experts from the equipment manufacturer and were the result of multiple trials designed to understand the complexity of the analysis of gypsum based systems.
Replicating the real board-line
While Henry Company has a test method that better mimics slurry manufactured on a board line, the short time frame in which the stucco gets mixed with the other components of the wallboard process in a pin mixer has not been duplicated. In full scale manufacturing, all components are mixed into an homogeneous slurry that is then deposited on the facing material in less than five seconds, whereas early efforts using a Ford cup or slump measurement in the laboratory could take several minutes.
During the evolution of this project, mixing the slurry constituents took at least 30 - 45 seconds and then the material was placed into the rheometer and tested. Unfortunately, even in this relatively shorter timeframe, much of the interaction of the gypsum particles with the additives had already begun and critical information leading to the understanding of why these dispersants may or may not work was lost.
With many process refinements, this mixing step has been reduced to 10-15 seconds before the slurry was introduced into the rheometer for evaluation. While this event horizon is approaching the 3-5 seconds of a wallboard line, it is still not as efficient as a production board line, so laboratory work has not been able to fingerprint the process from the start. However the procedure improvements have significantly improved the understanding of the component interactions in terms of wet-out and mixing. This has allowed the identification of differences between the additives and has allowed Henry to relate chemical signatures, structures and interactions to slurry performance.
Laboratory results and discussion
Figure 6 is a summary of peak-hold shear scans showing viscosity reduction of the test stucco slurry as a function of traditional additives (sulphonates) and the experimental additives being developed. There is a wide variation in their effect and in the case of the experimental additives, variation is seen due to chemical differences.
The lowest mean viscosity of the group was the Henry C additive, which also had a mean set time value that appeared to show little retarding effect on set (as demonstrated by Figure 7). This additive may have a particular application niche in set-limited board lines that currently use naphthalene sulphonates or are forced to overuse lignosulphonates.
Stucco variability can vary tremendously. Figure 8 shows the tremendous difference in dispersant effects not only between natural and synthetic stucco, but between two natural stuccos. The first stucco is a natural product and mined in the western north central part of the US2. The second set of data is based on natural stucco mined in the Mid-Atlantic region. The last set of data is based on synthetic stucco also from the Mid-Atlantic region.
The natural stucco from the West North Central Region shows the best mean viscosity value with the Henry C dispersant, but the naphthalene sulphonate and lignosulphonate dispersant means show that the naphthalene sulphonate works better than the lignosulphonate for the West North Central Region natural stucco, but worse than the lignosulphonate for the Mid-Atlantic region natural stucco.
Likewise, the mean viscosity value for the Henry B dispersant shows that it works better than the other dispersants for the Mid-Atlantic region synthetic stucco and the Henry C dispersant only works about the same as the naphthalene sulphonate and lignosulphonate. This exemplifies the sensitivity of the dispersant to the different chemical and/or physical properties of stuccos used by different wallboard manufacturing plants.
Three water : stucco ratios were examined as shown in Figure 9. A ratio of 0.7 was used because it was closer to the lower water : stucco ratios used by manufacturing plants, but without the rapid high viscosity build that interacts with rheometer equipment and useful measuring cycles. The higher water : stucco ratio caused more spinout of slurry, which made viscosity measurements inaccurate.
At the lower water : stucco ratios, which would be more desirable for wallboard plant formulations, the distinction between the dispersants increased, with the Henry additives showing improved performance over the lignosulphonate and the naphthalene sulphonate as shown in the bottom graph in Figure 10 at a water stucco level of 0.6 with 0.075g retarder (1 drop) used in each mix.
This reproducible trend is being evaluated at lower water stucco ratios to determine the practical limit as the additive chemistry is modified. A significant note is that in practical use (also duplicated in the laboratory) as the level of classical sulphonates is increased, there is a maximum amount of water that can be removed before the effect asymptotes and further addition shows no effect. The Henry additives are effective to much higher levels with associated higher water removal at equivalent rheology.
Field trial results and discussion
A limited set of field trials were conducted based on an optimised additive package developed in the laboratory using a customer's stucco. Figure 11 shows water reduction and associated kiln gas reduction as a function of additive dosage. Owing to logistics at the customer's plant, this trial was conducted using an additive package developed for incorporation in Henry's line of moisture-resistant wax additives and accounts for the higher order of magnitude addition rate that is reported.
During this trial the maximum amount of additive used was 92lb/msf for a gas reduction of 125cfm/msf. A gas volume reduction equates to an energy saving and therefore a corresponding cost saving. It was noted during the trial that further water could have been removed but the customer's goals had been satisfied and the trial was concluded.
Additional trials have been conducted in non-WR (water-resistant) board manufacturing where similar effects have been measured. These trials were not merely experiments because the different customers had different targets. For example, a review of a field trial conducted in the first quarter of 2010 resulted in the customer raising line speed and lowering kiln temperatures (lower gas usage) to a specific target. As a consequence the technology was not fully evaluated with the customer's stucco, slurry chemistry and process parameters, (i.e.: adding more additive and going to higher line speeds and/or lower kiln temperatures).
The results of this trial are shown in Figure 12. Samples 1 and 2 were plotted to establish a baseline with naphthalene sulphonate. Sample 3 represents the first addition of Henry additive and sample 4 the last and highest amount of additive, which had no deleterious effects on processing. As the additive concentration is increased, the trends show:
- A reduction in board density enabling higher performance or lighter boards,
- Significant removal of slurry water enabling higher line speeds and thus lower gas consumption,
- An increase in board strength, (MOR and nail pull).
Of further significance, based on laboratory results, the positive trends should continue with higher additive incorporation. This is being pursued in the laboratory and further customer trials.
A desire of many wallboard manufacturers is to decrease board density or weight while still maintaining the board's mechanical properties. Many strive for a more open core that provides a strong geometric gypsum core matrix through the distribution and size of the pores in the core.
Figure 13 shows a side-by-side light micrograph comparison of boards containing the Henry additive that has successfully 'opened up' the core structure with increased addition. This effect was accomplished using an optimised additive formulation based on stucco. Additional testing is currently being conducted to elucidate the cause and effect of these additives and their relation to board strength.
A common way to compare strength effects of wallboard samples with different additives is to use a static bending procedure outlined in ASTM C473.115. To estimate the toughness imparted to panel, the area under the stress-strain curve from ASTM C473.11 can be used. Figure 14 shows the stress-strain curves during flexural testing of a non-WR board with and without Henry's additive. As is shown, although the peak strength is comparable, the area, or work under the curve is substantially greater in the board with additive.
A more direct way to estimate toughness that simulates the effect of the system of framing members as well as the wallboard panel is ASTM C1629/C6, commonly know as hard body impact testing.
As Henry additives have been used to reduce water from wallboard, the hard body impact and bending strength performance tends to increase as shown in Figure 15 for hard body impact data, even with decreased board densities.
This effect will be evaluated further in future laboratory and field work.
Summary and conclusions
In this study we have evaluated the rheological relationship of surface active chemistry (SAC) to components normally used in wallboard manufacturing. Of the materials evaluated - stucco, dispersants, retarders and SACs, the effect of retarder is minor. Laboratory efforts with individual stuccos will be able to define the additive design space for individual plants, but because each wallboard manufacturing plant and process are different, iterative field trials will be needed to further refine the chemistry.
The Henry testing protocol developed in the laboratory more closely mimics the real-world process than previous procedures and allows a quicker targeting of the design space for chemical additives to be evaluated on the wallboard line.
Based on this study, a new proprietary class of additives has been developed both for WR and normal board that allows the wallboard manufacturer to reduce the amount of water necessary to achieve the proper stucco slurry rheology in the pin mixer and on the forming table, without adversely affecting the total process or the finished board properties. The incorporation of these additives has also demonstrated improved raw material utilisation as well as allowed better processing economics based on energy and line efficiencies.
Additional Henry chemistries have been shown to improve board strength allowing a potential reduction in density with a concurrent reduction in manufacturing cost.
Further efforts are underway at the Henry Company research and development laboratory to improve and extend this technology further into wallboard manufacturing as well as other applications.
References
1. TA Instruments. AR2000 Advanced Rheometer. TA Instruments Inc. New Castle, DE, USA. www.tainstruments.com, 2010.
2. PRo Publications International Ltd., 'Global Gypsum Directory.' PRo Publications International Ltd. Epsom, UK, 2007.
3. ASTM. 'Standard Test Methods for Physical Testing of Gypsum, Gypsum Plasters and Gypsum Concrete. Section 9. Normal Consistency of Gypsum Concrete.' ASTM International. West Conshohocken, PA, USA 2004.
4. Minitab. 'Minitab®15.1.1.0 Minitab Inc. State College, Pennsylvania,' USA, 2009.
5. ASTM. 'Standard Test Methods for Physical Testing of Gypsum Panel Products. Method B Constant Rate of Cross Head Speed. Section 11. Flexural Strength.' ASTM International. West Conshohocken, PA, USA.
6. ASTM. 2006. 'Standard Classification for Abuse-Resistant Nondecorated Interior Gypsum Panel Products and Fiber-Reinforced Cement Panels. Section 6.4 Annex A1. Hard Body Impact Test.' ASTM International. West Conshohocken, PA, USA.
